Human condition evaluation system, computer program, and computer-readable record medium

Abstract

A human condition evaluation system, of which bio-signal analysis means includes a bio-signal peak value detection means for detecting a peak value for each cycle of an original waveform of a bio-signal data, and a power value calculation means for calculating a difference between the peak value at an upper limit side and the peak value at a lower limit side for every prescribed time range using respective peak values obtained by the bio-signal peak value detection means to thereby set the difference as a power value, is structured to determine a predictive signal for falling asleep from an active state into a sleep state. Different from conventional human condition evalution based simply on a Lyapunov exponent as an indicator, a new human condition evaluation system based on a functional status of an energy generation system of the living body is provided.

1. A human condition evaluation system comprising:
 a bio-signal measurement instrument effective to measure a bio-signal of a human and to gather bio-signal data based on said measurements, 
 a bio-signal analyzer effective said bio-signal data;
 wherein said bio-signal analyzer includes; 
 a bio-signal peak value detector effective to detect a maximum and minimum peak value for each cycle of an original waveform of the bio-signal data, 
 a power value calculator effective to calculate a difference between the maximum peak value of said waveform and the minimum peak value of said waveform for each prescribed time range using respective peak values obtained by said bio-signal peak value detector to thereby set the difference as a power value, 
 a gradient calculator effective to obtain a gradient of the power value with regard to time base in a certain time range by performing slide calculation a prescribed number of times at a prescribed overlap rate with regard to the prescribed time, and 
 a comparative determinator effective to comparatively determine whether a sudden drop state of the gradient of the power value exists or not in time-series change in gradient of the power value obtained by performing the slide calculation to thereby determine the time range in which the sudden drop state appears to be a predictive signal for falling asleep from an active state into a sleep state. 
 
 
   
   
     2. The human condition evaluation system according to  claim 1 , wherein said bio-signal analyzer includes a Lyapunov exponent calculator effective to calculate a Lyapunov exponent by performing chaos theory analysis on the bio-signal data, and a Lyapunov exponent peak value detector effective to detect a peak value for each cycle of a time-series change waveform of the Lyapunov exponent calculated by said Lyapunov exponent calculator,
 wherein said gradient calculator includes a means for obtaining, in addition to of the power value, a gradient of respective peak values of the Lyapunov exponent with regard to time base in a certain time range obtained by said Lyapunov exponent peak value detector, and 
 wherein said comparative determinator comparatively determines whether a sudden drop state exists or not in at least one of the time-series changes in the gradients of the power value and the Lyapunov exponent obtained by performing the slide calculation to determine a range in which the sudden drop state appears to be the predictive signal for falling asleep from the active state into the sleep state. 
 
   
   
     3. The human condition evaluation system according to  claim 2 , wherein said comparative determinator compares the time-series changes in the gradients of the power value and the Lyapunov exponent which are obtained by performing the slide calculation with said gradient calculation means, and determines whether the gradients of the power value and the Lyapunov exponent are in opposite phases with each other or not before or in the range in which the sudden drop appears in the gradient of the power value or in the gradient of the Lyapunov exponent, and in the case where the sudden drop in the gradient of the power value or in the gradient of the Lyapunov exponent appears together with the opposite phase, said comparative determinator determines the range to be the predictive signal for falling asleep from the active state into the sleep state. 
   
   
     4. The human condition evaluation system according to  claim 1 , wherein said comparative determinator is further effective to determine that the transition into the sleep state is made when each time-series change in the gradient of the power value or of the Lyapunov exponent appears at a low amplitude on the whole, after the sudden drop appears in the gradient of the power value or in the gradient of the Lyapunov exponent. 
   
   
     5. The human condition evaluation system according to  claim 1 , wherein said bio-signal peak value detector is further effective to carry out differentiation for smoothing of the bio-signal data to identify a range in the vicinity of a differential waveform gradient at zero degrees to thereby detect the peak value from the original waveform corresponding to the range identified. 
   
   
     6. The human condition evaluation system according to  claim 2 , wherein said Lyapunov exponent peak value detector is further effective to carry out differentiation for smoothing of the Lyapunov exponent to identify a range in the vicinity of a differential waveform gradient at zero degrees to thereby detect the peak value from the original waveform corresponding to the range identified. 
   
   
     7. The human condition evaluation system according to  claim 1 , wherein said power value calculator is further effective to calculate as a power value, a difference between an average maximum peak value and an average minimum peak value both in a certain time range of the bio-signal data. 
   
   
     8. The human condition evaluation system according to  claim 1 , wherein the gradient of the power value or of the gradient of the Lyapunov exponent calculated by said gradient calculator is a value obtained by a least-squares method. 
   
   
     9. The human condition evaluation system according to  claim 1 , further comprising a frequency analyzer effective to perform frequency analysis on the gradient of the power value or the gradient of the Lyapunov exponent to determine a human condition. 
   
   
     10. The human condition evaluation system according to  claim 1 , further comprising an output means for actuating an awakening means for awakening a person when said comparative determinator detects the predictive signal for falling asleep from the active state to the sleep state. 
   
   
     11. A computer program, embodied on a computer readable medium, for letting a computer execute a process of evaluating a human condition by analyzing a bio-signal data detected by a bio-signal measurement in a bio-signal of a human, the process comprising:
 detecting a peak value for each cycle of an original waveform of the bio-signal data, 
 calculating a difference between the peak value at an upper limit side and the peak value at a lower limit side for each prescribed time range using respective peak values obtained by said bio-signal peak value detection step to thereby set the difference as a power value, 
 obtaining a gradient of the power value with regard to time base in a certain time range by performing slide calculation a prescribed number of times at a prescribed overlap rate with regard to the prescribed time, and 
 comparatively determining whether a sudden drop state of the gradient of the power value exists or not in time-series change in gradient of the power value obtained by performing the slide calculation to thereby determine the time range in which the sudden drop state appears to be a predictive signal for falling asleep from an active state into a sleep state. 
 
   
   
     12. The computer program according to  claim 11 , further comprising:
 calculating a Lyapunov exponent by performing chaos theory analysis on the bio-signal data, and 
 detecting a peak value for each cycle of a time-series change waveform of the Lyapunov exponent calculated by said Lyapunov exponent calculation step, 
 wherein said gradient calculation step includes
 obtaining, in addition to the gradient of the power value, a gradient of respective peak values of the Lyapunov exponent with regard to time base in a certain time range obtained by said Lyapunov exponent peak value detection step, and 
 
 wherein said comparative determination step comparatively determines whether a sudden drop state exists or not in at least one of the time-series changes in the gradients of the power value and the Lyapunov exponent obtained by performing the slide calculation to determine a range in which the sudden drop state appears to be the predictive signal for falling asleep from the active state into the sleep state. 
 
   
   
     13. The computer program according to  claim 12 , wherein said comparative determination step compares the time-series changes in the gradients of the power value and the Lyapunov exponent which are obtained by performing the slide calculation of said gradient calculation step, and determines whether the gradients of the power value and the Lyapunov exponent are in opposite phases with each other or not before or in the range in which the sudden drop appears in the gradient of the power value or in the gradient of the Lyapunov exponent, and in the case where the sudden drop in the gradient of the power value or in the gradient of the Lyapunov exponent appears together with the opposite phase, said comparative determination step includes determination of the range to be the predictive signal for falling asleep from the active state into the sleep state. 
   
   
     14. The computer program according to  claim 11 , wherein said comparative determination step includes determining that the transition into the sleep state is made when each time-series change in the gradient of the power value or of the Lyapunov exponent appears at a low amplitude on the whole, after the sudden drop appears in the gradient of the power value or in the gradient Lyapunov exponent. 
   
   
     15. The computer program according to  claim 11 , further comprising a step of performing frequency analysis on the gradient of the power value or the gradient of the Lyapunov exponent to determine the state of the human condition. 
   
   
     16. The computer program according to  claim 11 , further comprising actuating an awakening means for awakening a person when the predictive signal for falling asleep from the active state to the sleep state is detected by said comparative determination step.